The Augmentation Divide

AI's productivity gains flow to those already positioned to use them, and in an unequal society that turns a powerful tool into a force for stratification.

Sign up for my Substack daily AI newsletter here.

See my AI Training course portfolio for corporate Business Leaders here.

Article link: https://johanosteyn.substack.com/publish/post/204267550?r=73gqa&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true

Follow me on LinkedIn: https://www.linkedin.com/in/johanosteyn/

We were promised that AI would level the field. Every junior employee would have the leverage of a senior analyst, every small firm the reach of a large one. The early evidence points the other way. The gains are real, but they are flowing to those already positioned to capture them, and in a society as unequal as ours, a technology that rewards existing advantage is not a leveller. It is an accelerant.

CONTEXT AND BACKGROUND

Start with who actually benefits. Academic studies compiled by Goldman Sachs imply that generative AI delivers a 23 percent average uplift to productivity, but those gains flow disproportionately to workers senior enough to wield them, not to the staff whose core value was executing the tasks AI now automates. The same pattern holds at the level of the firm. A study of nearly 22,000 companies found that well-resourced, fast-growing adopters are pulling further ahead, while those that merely run pilots without sustained investment see no gains at all, opening a widening divide between the firms with the capital and talent to exploit AI and those stuck experimenting. Advantage, in other words, is behaving like capital: it accrues fastest to those who already hold it.

INSIGHT AND ANALYSIS

This is the augmentation divide, and its logic is compounding. The senior professional becomes a kind of super-employee, their judgement amplified, while the junior whose tasks are automated is left with less to offer. The well-capitalised firm turns AI into growth, while the under-resourced one falls further behind. An authoritative international assessment now warns that general-purpose AI could widen income and wealth inequality both within and between countries, shifting earnings from labour toward the owners of capital and the firms that develop or deploy the technology, with rich, digitally-equipped economies capturing the benefits faster than poorer ones.

I have written before about the rise of a digital oligarchy, a handful of foreign firms controlling the AI stack from chips to cloud to applications, and why South Africa, with its fragile education system and deep inequality, cannot treat this as a niche concern.

The honest counter-case deserves a hearing. The International Monetary Fund’s managing director, Kristalina Georgieva, has argued that although AI’s gains accrue first to high earners, rising productivity at the top can spill over to lift lower-wage workers too, a genuine reason for hope. The difficulty is that spillover assumes conditions a developing economy may lack: broad digital access, the skills to use the tools, and the capital to deploy them. Where those are scarce, the gains concentrate and the spillover thins. South Africa, with youth unemployment near 44 percent and a wide digital divide, is precisely the kind of economy where the optimistic mechanism is least likely to operate on its own.

IMPLICATIONS

For a country already stratified by income and education, this adds a new and sharper axis of division between the augmented and the rest. And the frontier is moving. Researchers at Yale recently demonstrated a brain-computer interface that let people control software by thought in under an hour, a signal, however early, that augmentation may eventually move from the tools we use toward capabilities embedded in the person. That future is not here, and should not be overstated, but the direction matters: an advantage you can buy is one a wealthy minority can buy first, and a cognitive enhancement is far harder to redistribute than a laptop.

CLOSING TAKEAWAY

The point is not to refuse the technology, which can genuinely lift productivity and create opportunity. It is to refuse the assumption that its benefits will distribute themselves fairly. They will not, unless access, skills and infrastructure are treated as deliberate policy goals rather than afterthoughts. The augmentation divide is not a distant risk but a present one, visible already in payrolls and balance sheets. For South African leaders and policymakers, the task is to widen access to the advantage before it hardens into yet another structural barrier, because a gap that compounds is far easier to close early than late.

Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net